A general guide for humans building AI rule systems, and for AI agents learning how to read and apply them. Covers: Cursor, Claude Code, Gemini CLI, Google Antigravity, OpenCode, OpenAI Codex, GitHub Copilot.

A prompt tells an AI what to do once.

A pattern for building personal knowledge bases using LLMs.

This is an idea file, it is designed to be copy pasted to your own LLM Agent (e.g. OpenAI Codex, Claude Code, OpenCode / Pi, or etc.). Its goal is to communicate the high level idea, but your agent will build out the specifics in collaboration with you.

Most people's experience with LLMs and documents looks like RAG: you upload a collection of files, the LLM retrieves relevant chunks at query time, and generates an answer. This works, but the LLM is rediscovering knowledge from scratch on every question. There's no accumulation. Ask a subtle question that requires synthesizing five documents, and the LLM has to find and piece together the relevant fragments every time. Nothing is built up. NotebookLM, ChatGPT file uploads, and most RAG systems work this way.

updated 2024-09-20

Version 2409? (Build 16.0.17928.20148)

updated 2023-07-14

Version 2304? (Build 16.0.16327.20264)

Create a production-quality Hermes skill named openrouter-expert that makes you excellent at building with OpenRouter and the OpenRouter SDKs.

This is not a notes file. It must be an agent-optimized resolver skill: compact, durable, triggerable, and designed to look up the right current docs before making claims or writing code.

Source material to read first:

1. OpenRouter docs index: https://openrouter.ai/docs/llms.txt
2. Full OpenRouter docs when needed for important details: https://openrouter.ai/docs/llms-full.txt
3. Skill creation best practices: https://agentskills.io/skill-creation/best-practices.md
4. Skill description guidance: https://agentskills.io/skill-creation/optimizing-descriptions.md

The way you approach solving a coding problem—how you think, communicate, and refine—is often more valuable than simply arriving at the right answer. Use ChatGPT as a supportive partner as you work incrementally through the six steps of the UMPIRE framework:

Understand → Match → Plan → Implement → Review → Evaluate